The subject matter discussed in the background section should not be assumed to be prior art merely as a result of its mention in the background section. Similarly, a problem mentioned in the background section or associated with the subject matter of the background section should not be assumed to have been previously recognized in the prior art. The subject matter in the background section merely represents different approaches, which in and of themselves may also be inventions.
Pattern verification, such as in a context of registration of a trusted fingerprint into a trusted memory of a fingerprint verification system, requires a sufficiently accurate and precise image of the trusted fingerprint (how accurate and how precise are dependent upon design considerations of a particular embodiment). This image may be obtained directly from an imaging sensor. This may be suitable for large area imaging sensors. However, for many applications, the trusted image is reconstituted from image portions of the trusted fingerprint taken from a sensor having a sensing area smaller than the final image. For many mobile applications and electronic platforms, such as smartphones and the like, the use of the small sensor provides a smaller form factor having a lower cost of goods which is important to implementation and adoption. However use of the smaller sensor impacts registration and user experiences during verification.
Any mobile electronic platform having a fingerprint verification system typically includes a registration system that establishes and writes one or more trusted images into a trusted database/memory. Test fingerprints are compared against the trusted image(s) using a matching engine that matches features of interest from the test fingerprint against features of interest from the trusted image(s). A quality of the test image and of the trusted images impact the user experience.
Whatever the size of the imaging sensor, the trusted image may be derived from a single impression from a single finger, or derived from a composition of multiple impressions. Each solution offers different challenges.
It is technically possible to rely on a single impression of a single finger using a small area sensor for registration. However, during use, the user must not only recall which portion of which finger was registered, but each verification attempt relies on the user reimaging this same portion of the same finger sufficiently close that it may be matched. As the imaging sensor becomes smaller, reliance on a single impression can become very frustrating for the user.
Many systems using a small area sensor therefore register a single finger using multiple impressions. An initial impression of an image portion always provides completely unique information of the larger image to be reconstituted. However, subsequent impressions have varying usefulness depending upon whether the user provides an impression that partially overlaps and partially presents new image data. The user has difficulties in placing a finger for multiple impressions to ensure only unique partial overlapping data is presented with each impression, particularly over a breadth of an entire fingerprint pattern. It is too easy for a user to provide subsequent images with too much overlap or with too little or no overlap. A user does not know when or if they have provided enough sufficiently overlapping image impressions that map enough of the total fingerprint to provide an end product that provides the user with an easy-to-use verification system.
From a perspective of a user, this can become very challenging as the sensor imaging area decreases. Registration systems do not provide feedback to the user of a status of the finger area mapping process. An accuracy of the registration process is degraded when the set of image portions are unable to produce the needed larger image.
Existing systems employ stitching systems that create a bigger image from ridge map information of the individual image portions. Such a system adds increased complexity to a registration system that, as further explained herein, already includes resources sufficient to map or reconstitute a single image from multiple images provided using multiple impressions of an image source.
A simple system may assume, without checking, that each impression from a multi-impression set of images are all unique qualifying images from a single finger. Systems that rely on a preset number of unqualified image impressions produce a varying quality for the final image to be used in the trusted database. Subsequent verification based upon these registrations results in a wide range of user experience in a quality of the verification process. Some users that are unfamiliar with registration may use different fingers in a registration system designed for a single finger registration process. Such image portions can also degrade the registration process in many different ways.
Systems for processing image portions of a pattern may sometimes be desirably implemented using limited computing and memory resources. As an active area of a sensor decreases, a greater number of image portions are needed to cover any particular area of the pattern. Managing and processing multiple portions can sometimes negatively impact performance, particularly if other systems and methods may be implemented to address this impact.
What is needed is a system and method for efficiently reconstructing a pattern, such as a fingerprint, from a set of multiple impressions of portions of that pattern.